Population Markov models for the analysis of public health policies

Risk factors, e.g. smoke and inactivity, affect quality of life and life expectation of people who are exposed to such risks. Prevention measures may impact on the exposition to risk factor, and therefore on their negative effects. In this thesis we build a model that simulate the evolution of a population subject to a risk factor and analyse the effects of public health policies on diseases related to the risk factor. In particular, the smoking risk factor and 4 diseases that are strongly correlated with smoke are taken into account. To measure the impact of the 4 diseases, we use two indicators of the disability burden induced by a disease, i.e. years of life lost due to premature death due to the disease and years lived with the disease. To describe the evolution of the population, we use Markov chains. We construct the initial population and the transition matrix of the single individual, based on italian data, in order to simulate as realistic as possible the italian scenario. Finally, we observe the effect generated by the implementation of a prevention policy that acts directly on the smoke prevalence in the initial population. To evaluate the effectiveness of the policy, we compare the statistical indicators obtained by simulating the actual population, with the ones obtained by implementing the prevention policy.